Fama-French Short-term Adjustment Model In The Study Of Applicability Of Chinese Stock

It is always the main research direction of investors to disassemble and study the factors that affect stock returns in order to obtain excess returns.In 1992,Fama-French proposes three factors that affect the investment strategy(return rate of the securities market yields on the slope coefficient of the regression)model.Since then,based on the Fama-French three factor model,the domestic and foreign scholars have widely discussed,extended,applied and validated the two aspects of the model in terms of market effectiveness and model setting.First of all,although the validity of the Fama-Frenchmodel has been preliminarily certified,there are still differences between the scholars at home and abroad on the basis of the determination of the three factor or the four factor.Second,the accuracy of the Fama-French model and its ability to increase model interpretation by adding new factors are widely divergent.Scholars at home and abroad have increased the explanatory power of the model from different levels,such as price earnings ratio,liquidity factor,etc.,and fully demonstrated the possibility of improving the accuracy of the model.Third,China's financial market started late,the regulatory system is not perfect,as well as national conditions,consumption,investment,habits and other aspects of the difference,quantitative technology localization research is the key to the future development of quantitative funds.Only by designing a quantitative model that meets the domestic market environment and achieves excellent results can investors be identified.Finally,both the Fama-French three factor model and the four factor model provide the long-term validity of the market.The existing scholars' research is also based on the long-term validity of the supplement and further fitting,and the short-term fluctuations and other factors have not been further disassembled and analyzed.That is,based on the characteristics of the data,the cointegration test of each variable is used to eliminate the nonstationarity of the time series,so as to establish the error correction model.After that,the model fitting result is predicted and verified according to the two period of formation and holding period,so as to realize the long-term fitting and short-term correction of China's Shanghai and Shenzhen stock markets.After a large number of actual data tests,the quantitative model of stock selection Fama-French-ECM model has strong adaptability in our country,which shows the following aspects:First,the model has a good fitting effect.Second,in addition to the secondary market to book company,the other two factors combination of firm size,book to market ratio has a strong ability to explain:The company size is negatively related to the excess return,and the book market value is positively related to the excess return.Third,Fama-French-ECM model improves the fitting degree of the original model.Based on the data characteristics,the cointegration test of the variables in the model eliminates the non-stationary nature of the time series caused by a large number of uncertainties;The model fixes and clears a substantial gap between the actual value of each portfolio return and the forecast value,After the short-term adjustment model was used in 2015,the accuracy rate increased by 9.75 percentage points on average,and the accuracy rate was over 95%in the first three quarters.Fourth,various parameters of the original model is not effective,which are selected according to the different market performance.In Fama-French model,the constant term,medium market ratio variable coefficients were not through hypothesis test;In Fama-French-ECM model,constant term of large scale low book to market value,small scale high book to market value,small scale and low book to market value,the book to market ratio of large scale medium book to market value,the difference item of book to market ratio and firm size factors of small scale medium book to market value were not breakdown by hypothesis testing,we excluded these items for the explanatory ability of the model,finally obtained the simplified model.Finally,according to the two periods of formation and holding period,the model fitting results are predicted and verified.In the short run,the change of the explanatory variables is determined by the stable long-term trend and the short-term fluctuation.In a short time,the system deviates from the equilibrium state directly affects the volatility amplitude.In the long run,co integration relation plays the role of gravity line,and can transform non-equilibrium state into equilibrium state,it can better fit China's CSI 300 stock market,analyze and judge the expected rate of return better,achieve the academic discussion and practical application.